Email Address:

Lost your password?

This is the legacy website; please use the new website.

Serviceman's Log

When the going gets tough.

By the Serviceman

Items Covered This Month
  • Sony TC-366 reel-to-reel tape recorders.
  • Sony KV-2184SA (GP-1A) TV set.
  • AWA 65850 S/T TV set.
  • Grundig ST70-755 TOP/LOG TV set (CUC 6360 chassis).
  • AWA C5105 TV set.

There are times when I swear blind that I will never touch any equipment that's over 10 years old again. That's because of all the warranty implications when something else inevitably fails and because you often cannot get a critical part or it is prohibitively expensive.

Click for larger image

But then I reach a quiet business patch and in a panic I accept not one but two 20-year old units for repair. Yes, I regret to say that muggins was guilty of this when two unrelated people brought in Sony TC-366 reel-to-reel tape recorders.

These Sony TC-366s are nicely built 3-speed machines. They have one motor, their fastest speed being 7.5ips (inches per second), and all the tape action is controlled via a large single knob. In both instances, the symptoms were no or little play torque but each fault was different.

Removing the covers involved unscrewing five Philips screws, undoing two special poles which also hold the head cover, and removing the Pause/Instant Stop switch knob. This round knob is about 1mm larger that the slot it is meant to go through when removing the tape deck escutcheon. It is screwed onto its shaft and one has to manually unscrew it in order to release the escutcheon from the machine.

Unfortunately, some 20 years later, this knob refused to unscrew, so I had to gingerly widen the slot with a round file. I really don't know why they couldn't have made it pass through when it was manufactured!

When I could finally see inside the mechanism, it was obvious that the large play idler wasn't moving to engage the motor and capstan shafts. Lubricating the bearings of the associated levers solved this problem on one of the machines.

The second was more complicated. The motor wasn't rotating at all until I gave it a little nudge. But although I could get it going, it didn't have much get up and go about it. In fact, it kept slipping, especially in fast forward, and was also getting very hot.

As a matter of interest, this particular machine had been imported from the USA and was connected to our 240V mains by a stepdown transformer. The machine was rated at 117V 60Hz and the AC motor at 100V 50/60Hz.

The 1.5μF starter capacitor (PN 117-034) looked to be the main culprit. As luck would have it, when I rummaged through an old box of capacitors I found not one but two original replacements still sealed in plastic bags. The only difference was that they had an additional 0.5μF terminal on them as well (PN 117-036).

This fixed the problem immediately but not being a rocket scientist when it comes to AC motors, I wasn't really sure whether to simply connect the 1.5μF section of the replacement capacitor or connect the 1.5μF and 0.5μF sections together to make 2μF for 50Hz (as opposed to 1.5μF for 60Hz).

Well, I initially left it at 1.5μF and put the machine on soak test. Three hours later, I revisited it and was horrified to find that the motor was nearly too hot to touch and had stopped. The replacement capacitor had failed and I had only one left.

This time I connected the two sections in parallel, lubricated the motor and soak tested it again. Fortunately, this one held and, hopefully, will continue to do so. If not, it means ordering in a very expensive industrial one.

Anyway, I finished up with two satisfied customers, so my gamble with these 20-year old machines paid off.

An interesting Sony

An interesting TV set came in the other day. It was a 1990 Sony KV-2184SA (GP-1A). So what is interesting about that? Well, the "SA" designation in the model number (rather than AS) indicated that it was from South Africa. In addition, it had a built-in FM radio on the lefthand side of the cabinet for simulcast transmissions.

This surprised me because I thought the whole purpose of simulcast transmissions was to receive stereo and yet this was fed into the monaural TV sound system and only one small elliptical loudspeaker. So I really can't quite see the point; perhaps some ex-pat can enlighten me, as there is no instruction book or service manual available for this model in Australia.

The set was dead, with an IC link fuse (PS801) open circuit. This led me on a merry dance as I was using my old service manual for a KV-2184AS and there was an extra transistor circuit (Q622) that wasn't marked. This supposedly monitors the current through R621 and turns off the horizontal drive from pin 27 of the jungle IC (IC301, CXA-12135) by biasing pin 22.

First, I found the D608 R2M safety zener to be short circuit, which implied the 115V rail had risen high enough to destroy it. So I replaced this, along with IC601 (STR50115, B version). I then found that R621 (1Ω) had gone high, as well as R622 (470kΩ).

I was getting closer but the set gave a high-pitched squeal when I switched it on and the 115V rail was at 87V. Yet when connected via a 100W globe, with a clip lead between base and emitter of the horizontal output transistor, the voltage was spot on at 115V and there was no squealing.

When I re-enabled the horizontal output stage, I could get sound and a white raster. And when I subsequently replaced R851 1.2W (200V supply to the CRT), I could get a coloured picture, albeit small and distorted due to the low HT.

From this, I deduced that the horizontal output stage might possibly be drawing too much current. But it wasn't and it took a very long time to work out that R609, a 33Ω 3W resistor, was open circuit in the power supply and that this was part of a feedback circuit to pin 2 of IC601.

By now, I had a good picture and sound when I connected a video recorder to the AV inputs. However, that wasn't the case off air.

Click for larger image

The reason for that wasn't hard to figure out. South Africa uses the CCIR system with a 6MHz sound IF and with no Band I VHF (V-L). This meant removing the IF module IF201 (IFB-389SA) and replacing the three 6MHz ceramic filters with 5.5MHz units (CF01 for D5.5E, CF02 5FE5.5MB and CF03 T5.5B). The set now gave good sound and colour for all stations except for the ABC (Ch2, Band I).

Now all that was left to be done was to fit the extra band switching transistor (Q151, DTA114ES) - along with 6.8V zener D152 and C150 (1μF 25V) - and link it to pin 7 of microprocessor IC002 via a 2.2kΩ resistor (R098). I did all this but it still didn't work. The transistor would still only scan high-band VHF (ie, no Band I) and UHF.

Initially, I thought this was due to the microprocessor being a PCA84C640P/037 instead of PCA84C640P/016 as in the circuit diagram. However, this was just a red herring as I discovered that the Australian module also used the same "037" suffix and worked OK.

What I did discover was an extra diode (D005) fitted in the South African model that connected pin 22 to the vertical pulse going to Q005 from the vertical output stage. Removing this diode breaks the TV scan mode into VL, VH and U and the set could now be tuned like an Australian TV receiver.

Obviously, the easiest way to solve the above problem would have been to fit an Australian tuner and an the IF module as well (as they would be better matched). However, I hadn't scrapped any of these sets lately, so I couldn't scrounge the necessary parts and fitting new units would not have been cost-effective.

Finally, to select the simulcast FM radio, it is necessary to rotate the system switch inside the front control panel and then tune the knob on the side. The tuner is connected via CN302 and CN201 to the de-emphasis pin (pin 10) on the IF module and pin 7 of analog switch IC702.

AWA TV set

AWA is now used as a brand name on a series of TV sets imported from China. These sets are very similar to those of Teac, Akai and Masuda, and are probably produced by the same ONWA factory.

In fact, the 1993 66cm AWA 6850ST on my bench could easily have been mistaken for a Teac CT-M711 or a Masuda 28AV. And it was completely dead.

After removing what seemed like an unnecessary quantity of screws, the set was seen to consist of a motherboard, a Teletext board, a stereo decoder and a power supply. And a quick examination soon revealed that two capacitors in the power supply had spilled their guts.

I replaced the two culprits - namely C910 47μF 25V and C908 10μF 50V - before replacing the parts they had destroyed: Q909 (2SD1403), R944 (2.2Ω, 5W), C925 (1000μF, 25V), R916 (0.68Ω, 1W) and C925 (22μF, 160V). Judging by the soldering, I wasn't alone in going down this path. I also checked the HT voltage and set it to 143V at TP B+.

Intermittent flicker

The next problem was the picture which was intermittently flickering horizontally with colour problems. Sometimes the picture would tear and sometimes it would shift horizontally. It's not only a horrible fault to describe but was equally hard to diagnose and fix. And it was somehow temperature related, the set getting better when hot.

By trial and error, I discovered that the fault seemed to be located around IC102 (TDA8305A). This is predominantly an IF detector processor but also a jungle IC and is located underneath the stereo decoder board in a soldered-in metal cage. There was a lot of black silicone rubber all over this set which, like the brown glue before it, tends to conduct after many years.

Well, I had to start somewhere so I scraped off the black silicone rubber and replaced the IC. This didn't work, so I started replacing the capacitors. Fortunately, I struck gold with C127 (3.3μF), which is a decoupling capacitor for the AGC detector and sync demodulator.

At this time, I didn't have a remote control but fortunately I had a Remote Master RM0900 Universal remote control and tuned that into this set's command set. This was a good idea, as I found then that the remote control power on/standby function wasn't switching off.

This set used a relay for switching the 143V on and off to the rest of the set, so it was just a matter of tracing it back to the microprocessor. However, the fault turned out to be a lazy momentary contact on the main on/off switch (S512) that was sticking in the on position. A little lubrication soon fixed that and the set was back in operation.

Grundig day

I don't see many Grundig TV sets these days, probably due to the dealer network set up by Grundig in recent years. I normally try to avoid house calls but when Fred Schmidt called and gave me his address as being only a block away, I bent the rules and arranged to "have a look" at his 10-year old Grundig ST70-755 TOP/LOG (CUC 6360 chassis). The fault complained of was intermittent no picture and it was the "intermittent" bit that worried me.

When I arrived, the set gave no picture but a green raster with retrace lines. After it had been on for sometime, the picture flashed on.

I removed the back - there were only four screws to undo (other manufacturers please note) - and checked for faulty joints or anything obvious. The only thing I learnt was that by varying the CRT screen control SG/VG2 clockwise to give a very bright green raster and then slowly reversing it anticlockwise, the picture would return. However, it was too bright and too green and turning the brightness control down on the remote control still couldn't make the blacks black.

Basically, if the screen control is turned too far anticlockwise, the picture loses definition and gives poor colour contrast.

Click for larger image

This looked to me like the RGB video output stage was producing signals that were outside the optimal characteristic curve sections of the CRT guns. I told Fred that it had to go to the workshop and though it was a fine TV set, perhaps he ought to think about updating it considering its age. However, being rather short of cash, he decided to have the set fixed rather than dumped.

Back on the workshop bench, I first replaced the RGB video output IC (IC790, TEA5101A/D), which is a very complex FET amplifier. Initially, I fitted a generic IC from my local supplier but later, when I became more desperate, and at Grundig's suggestion, I ordered and fitted a genuine replacement but it made no difference either way.

This type of circuit is used in a lot of European-designed sets like Philips and the fault can quite often be caused by a low-emission tube. To prove this one way or another, I decided to test the Philips A66 EAK71X01 picture tube thoroughly. First, I connected my CRT analyser directly to the tube and, at 6.3V RMS, I measured an even 0.62mA emission from each gun, which is reasonably good.

Next, I measured the voltage across the CRT filaments with my RMS voltmeter and that too was spot on at 6.3V. Finally, I connected an external power supply to the heaters and wound it up to 8V to see if it made any difference but it didn't.

I now concentrated on the diodes and electrolytic capacitors on the CRT module (29305-022.48). Note that this differs from the one in the service manual (29305-022.44) - it has four more diodes fitted, as well as other modifications.

I tried replacing all the 1N4148 diodes with higher-rated BAV21 (used mostly as clamps) but that didn't help. I also measured all the voltages around the CRT socket but of course, the voltage for each colour was not consistent and besides, the service manual does not give voltages.

The thing I did notice was that the voltage drive for each colour from the colour decoder module was 4V for red and blue and 5V for green. I also measured the beam limiting SB voltage from the horizontal output transformer (A) on pin 1 of the module to be 16V and the SW voltage on the CRT RGB pin 1 plug to be 0.63V in the fault condition. I should also add that it was not possible to set up the screen control according to the alignment instructions, as the picture would be far too bright (in AV mode, there should be 172.5V on each cathode). The EHT was spot on.

Colour module

All this was very time consuming and all I had really achieved was to confirm that the CRT and socket were probably OK. So now I shifted my attention to the colour module - 29504-105.67 - which also differs from the 29504-105.56 module in the service manual.

The other drama with this double-sided module concerned the access to its surface-mount components. About all I could measure were the voltage supplies (12V, 8V) and the inputs and outputs. Fortunately, the service manual gave a pretty good explanation of the circuit.

The circuits I was interested in were mostly in IC5040 (TDA8376) and the 10 surface-mount transistors after it. I pulled the module out and using an ohmmeter, checked all the transistors to be OK. I then replaced all the electros and measured all the diodes. An exchange module (if available) would cost $227.25 trade and the TDA8376 $119.70 plus freight charges of $11.

I gambled on the latter and ordered it. A few days later, after replacing this 52-pin high-density IC, I discovered I had lost and blown all my money!

So far I wasn't doing very well. If only I had had another set to compare it with. As I had the necessary ICs in stock, I replaced all the other ICs on the Chroma Module but even that was a waste of time.

Finally, I decided to replace the surface mounted transistors, starting with CT5127 (Red), CT5128 (Blue) and CT5129 (Green) - all BC858s. Naturally, with my luck, these are no longer available, so I fitted BC858Cs. This fixed one problem - the green caste was gone and the grey scale was correct even if it was too bright.

I surfed the net and found one German site for TV and Video Service Repair Tips that had to be translated, suggesting I replace CT5066, another BC858. This I did and I noticed a small improvement. Thus encouraged, I moved along and replaced CT5060 as well. This finally produced an acceptable picture after realignment. These transistors are part of the beam limiting current circuit to pin 22 of IC5040 (3-4V).

I couldn't fault any of the PNP transistors I replaced but because of their size and position, it was almost impossible to hook up complicated test gear to test them. I have even had cases where the glue used to cement these components onto the board affected their characteristics.

Anyway, the picture was now quite acceptable but will Fred go ahead with my quote? How could I possibly guess at the likely cost beforehand?

And to finish off, here is another reader contribution. It is from A. B. of Glen Waverley, Victoria. He has titled it:

The perils of modern test gear

Mr Smith staggered up the path clutching an AWA C5105 TV set, moaning "I couldn't find the place". I resisted the temptation to ask if he had seen the sign out the front but in this game you soon learn that the customer has a good reason for whatever they say.

"Its dead, just stopped - is it the picture tube?" he puffed on. After reassuring him it was unlikely, I booked it in and shoved it onto the "look at" pile.

Click for larger image

I got to it a few days later. It was a 51cm AWA set of Chinese manufacture and only about 18 months old. The first thing you do in these sets is go straight for the power supply, in this case the electrolytic capacitors in the primary circuit (C614 and C615). Both measured faulty on the ESR meter and were promptly replaced. As well, the EHT circuit was checked for shorts and a DC resistance measurement to chassis showed no obvious faults.

The set was then powered up and I fully expected it to burst into life. Alas, it was not to be - the red standby LED was on but no amount of prodding at the usual user controls made any difference.

At this point, you start to wonder if the remote control is needed to start the set - most modern TV sets are switched on by firing a start-up pulse from the remote using the on/off button or a channel button. I worked on this assumption and turned the chassis upside down on the workbench to access the underside of the board.

It was soon established that the HT rail was at 110V - a figure I assumed to be OK on a set this size. After diligently studying the circuit, a 5V regulator and an 8V regulator were also found to be functioning. As well, there was HT on the collector of the horizontal output transistor and about 85V on the collector of the driver transistor.

However, a quick check with the CRO revealed that there was no drive waveform to the base of the driver transistor. Unfortunately, the relevant track soon disappeared into a mess of similar tracks wandering all over the board.

At this point, it was decided to obtain a circuit diagram. Unfortunately, when this turned up, it proved to be a 15th generation copy (or thereabouts), plus a booklet full of useless (in a practical sense) information on how to do alignments, etc.

The circuit was of limited help as all the figures were unreadable. However, I did manage to locate the jungle IC and I checked that this had supply volts. I also discovered that there was no horizontal drive waveform on the relevant output pin.

So what was going on? All the routine tests had been done to no avail. After staring at the board, I decided to measure all the high-wattage low-value resistors. All checked OK (or as near as could be judged given the useless circuit diagram) with the exception of R320, the dropping resistor to the horizontal driver circuit - it was reading about 6MΩ! However, the HT on the transistor measured OK at 85V!

Suddenly the penny dropped, and I mentally kicked myself around the workshop. I pulled the resistor out and replaced it with a 2.7kΩ 5W unit, switched on the power and the set came to life.

What happened?

So what had happened - easy, and if I had remembered basic theory I would have found it a lot quicker. What happened was this - at switch on, the start-up pulse gives enough energy to kick the set into life and all is well. In this case, because the set wasn't running, my high impedance digital multimeter was indeed reading a nearly correct voltage across the driver transistor because the thing was still stuck in standby mode, the transistor not drawing any current.

By contrast, an old, cheap analog meter with its higher loading would have shown a lower voltage at this point and the true fault would have been revealed at once!

Ah well, that's the problem with getting older - what was once routine is now new if you haven't seen it for awhile.

Share this Article: 

Privacy Policy  |  Advertise  |  Contact Us

Copyright © 1996-2019 Silicon Chip Publications Pty Ltd All Rights Reserved